Cloud Security Podcast by Google EP238 Google Lessons for Using AI Agents for Securing Our Enterprise
Aug 11, 2025
Dominik Swierad, Senior Product Manager for D&R AI and Sec-Gemini, shares his insights on implementing AI agents within Google's security framework. He discusses strategies for building trust among security professionals and identifies key use cases for AI integration. Swierad emphasizes the importance of high-quality data and structured methodologies for evaluating AI performance. He also highlights the evolving roles of security engineers amidst rising automation and raises concerns about how threat actors might exploit similar technologies.
AI Snips
Chapters
Books
Transcript
Episode notes
Start With Simple GenAI Chat
- Integrate a simple generative AI chat into security tooling so engineers can query their own data.
- Dominik Swierad says that hands-on exploration drove consistent monthly usage and built confidence for larger agent work.
Start Where You Have Good Data
- Choose initial use cases where you have many clear examples and representative data.
- Dominik Swierad warns that lacking good examples makes a task a poor starting point for AI.
Focus On Bottlenecks, Not Hype
- Prioritize agent projects that remove true operational bottlenecks instead of applying AI everywhere.
- Dominik Swierad stresses delivering value where teams can actually absorb and act on increased coverage.
